One-Way Versus Two-Way Videotex

(1)

NOT FOR QUOTATION WITHOUT PERMISSION OF THE AUTHOR

ONE-WAY YERSUS TWO-WAY VIDEOTEX

H.A. Maurer I. Sebestyen

March 1982 WP-02-30

Wmking Papers are interim reports on work of the International

Institute for Applied Systems Analysis and have received only limited review. Views or opinions expressed herein do not necessarily represent those of t h e Institute or of its National Member Organizations.

INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS 2361 Laxenburg, Austria

(2)

One-way and two-way videotex are often claimed to be more or less competing services, particularly if dedicated channels are used for the broadcast variety. In this paper we will try to dispel this notion. We will try to demonstrate that not only are the two services somewhat comple- mentary in nature but that videotex will increase its potential by choos- ing a balanced combination of the two services. We also study the likely market penetration of videotex based on the speed of penetration of other communications-oriented services in the past and draw some conclusions how they might influence the market penetration of videotex systems and services.

(3)

CONTENTS

1. INTRODUCTION

2 . CURRENT AND FUTURE VIDEOTEX TERMINAL TECHNOLOGY 3. CURRENT AND FUTURE 1WN (ONE-WAY NARROW

BAND) VIDEOTEX

4. T H E IMPACT O F 1WW (ONE-WAY WIDE BAND) VIDEOTEX

5. T H E ROLE O F 2W (TWO-WAY) VIDEOTEX 6. T H E COMBINATION O F 1W and 2W VIDEOTEX

7. SOME ASPECTS O F THE MARKET PENETRATION O F 1W and 2W VIDEOTEX

7.1 T e l e c o m m u n i c a t i o n I n f r a s t r u c t u r e f o r 1W and 2W V i d e o t e x

7 . 2 V i d e o t e x H a r d w a r e and S o f t w a r e System components

8. CONCLUSIONS REFERENCES

(4)

ONE-WAY YERSUS TWeWAY VIDEOTEX

H.A. Maurer and I. Sebestyen

1. INTRODUCTION

The basic idea of videotex is to add sufficient electronics to TV sets so that they can be used not only for receiving

TV

programs but also as terminals of a (potentially computer-supported) information system.

Around the end of the 1970s videotex systems came along in two varieties:

as b r o a d c a s t or one- w a y (1W) systems and as i n t e r a c t i v e or t w o - w a y ( 2 W ) systems. In the first case, a set of information p a g e s ("frames"), each identified by a number, is available. The user can select which of the information pages (usually containing textual information) he wants to look a t , usually by keying in the number of the desired page using his TV remote-control unit, TV set. In the second case, the user can not only retrieve information pages, but can also send information w b c h can be used for ordering, booking, communication, and many other interactive applications.

One-way videotex can be of two types. The narrow-band version (1WN videotex) is the older, "classical" version

--

first introduced as C e e f a z by t h e BBC and Oracle by the IBA in the UK. In the same category, Austria's T e l e t e z t was the first nationwide 1WN videotex service and has now over 100,000 participants. Other countries have since introduced 1WN videotex on a nationwide scale, such a s West Germany's V i d e o t e z t . In all these 1WN videotex systems some few hundred information pages are sent on a rotating basis on the same channel and are simultaneously mixed with an ordinary TV program (using the vertical and horizontal blanking intervals for transmission), w h c h has to be "demixed" by a so- called teletext decoder to prepare the relating flow of information frames

(5)

for selection. In this fashion l.ess than 10 frames of information c a n be transmitted per second. Hence, to avoid annoyingly long waiting times for a page specified by a user, only a moderate number of frames (some hundred) can be sent before retransmission has to s t a r t again.

The second version of !W videotex uses a fu.l!y "dedicated" TV chan-:

nei, usually available via cable TV (CATV) or, in countries with few TiJ pro- grams or empty channels, one could imagine a separate broadcast TV channel devoted to teletext. The basic idea of this wide-band (1WW) videotex is similar to t h a t of 1WN except t h a t much larger quantities of information can be sent within the same amount of time if a mcr;. power- ful, faster, and more sophisticated teletext decoder is used. On average, some 50,000 frames can be offered within a few seconds.

In contrast to the i W vidsotex, in the two-way (2W) version t h e user has a separate channel for communicating in the other direction. The most common implementation of t h s idea is to use switched public telephone lines for comn~unication in both directions. Thus the

TV

s e t acts as simple computer terminai which is hooked up -- via a telephone dialing line

--

t o the computer of t h e videotex center (which in t u r n may act as gateway to other "external" or "third-party" computers t o create a videotex network or, as they call it in Germany, Rechnerverbund).

Although the

TV

s e t as videotex terminal has all of the functions of a computer terminal, much of t h s potential (but in varying degrees), is not used in current videotex systems. This is partially due to very rudim-en- tary "keyboards" (often only the numeric keypad TV remote-control unit) available t o the user; to the lack of software in the videotex centers; to t h e rudimentary state of t h e videotex network (if available a t all) and to i h e "primitiveness" of the terminal. However, even in the simplest vari- hnts, the feedback option of 2W videotex does offer many possibilities not b a i l a b l e in 1W systems.

We will discuss some further special aspects of videotex systems required for comparing 1W and 2W systems in the sections to follow, but refer to studies such as [6], [15] and [18] for a broader overview. We plose this section by noting t h a t other terminal implementations of both

! W

and 2W videotex systems a r e also feasible and are or have been experi- mented with: e.g., using a full radio channel for 1W videotex, using CAW with feedback channel for 2W videotex, or both using CATV (in one direction) and telephones (in t h e other) for 2W videotex.

2.

CURRENT

AND FVWRlZ VIDEOTEX

TERMINAL

TECHNOLOGY

The most widely used types of videotex today a r e those based on early developments in UK (Ceefuz, Oracle and Prestel); these make use of very simple (i.e. "dumb") terminals with very little local processing and storage and (in t h e simplest version) only numeric keypads. In contrast with these terminals and some of their planned successors in Switzerland, Holland, and West Germany, for example, other countries have decided t o develop terminals with more local processing capability (Canada and the US), and alphanumeric keypads a s standard input devices (France).

Furthermore, some existing personal computers ( s u c h as t h e Apple II), have been made "videotex compatible" by adding appropriate interface

(6)

cards to the basic system.

With the falling price of integrated circuits we expect that future videotex terminals will tend to have more and more "intelligence" and alphanumeric keyboards a t ever-decreasing prices. Such local "intelligence" (which might put the future videotex terminal, or a version of it, somewhere between a videotex terminal and a personal computer of today) will make the use of videotex easier, will allow certain amount of decentralization and will open up the possibility for applications such as using telesoftware [ I 31. The availability of local storage and intelligence will also be particularly useful in connection with 1W videotex, especially in the wide-band version. (See Sections 3 and 4).

It is becoming increasingly evident that future videotex terminals will allow external attachments such as to tape recorders (already in use in Holland, for instance) and printers, to enable hard copies of frames of interest to be printed. Although one cannot expect that all ordinary households will: be equipped with printers in the foreseeable future,

"semi-local" printing devices (such as one printer for a whole apartment building) could become a reality within the 1980s and could increase the impact of videotex considerably. In addition, cheap printers for personal computers are already available. In the UK, for example, Sinclair offers to its XZ81 computer a small matrix printer for less than # 50.

3. CURRENT

AND

FUTlTRE lWN (ONE-WAY NARROW

BAND)

VIDEOTEX

Due to the fact that only some 100 frames (corresponding to less than t e n newspaper pages) can be sent in the usual 1 W N videotex systems within some 20 seconds, the applicability of 1WN videotex appears to be severely limited. Typically, the 100 pages of information sent in the early days of the Austrian service (early 1980) were a t most moderately useful.

In a small experiment, one of the authors found that after some initial high usage due to the novelty of the service, usage dropped to less than once per user per month. An analysis of the reasons for such very limited usage has revealed two major factors: one is limited (and possibly

"wrong") information presented, and the other is the cumbersome access procedure. To find a specific piece of information a number of index pages have to be looked up first, each look-up lasting between 0-20 seconds (which tend to seem like a n eternity!).

Much experimenting with t h e type of material and how to present it has improved 1WN videotex remarkably. The authors feel that further improvements along the same lines and using some new ideas suggested below will make it an interesting and, because of the favorable cost- benefit ratio, a permanent feature of W .

Some of the improvements of 1W videotex which have been implemented in a number of countries such as Austria are as follows.

Rather than sending all pages with the same frequency, one new idea is to send them a t widely different frequencies, and this can be used in two ways. One is t o send survey and index pages more frequently to allow quick access to routing information, as is done in t h e West German system, for example. The other is to group pages (which one can assume will be accessed by the user mostly either not a t all or else all of them) into

(7)

"magazines" where the pages are transmitted a t , say, 20 second intervals.

Thus, some 15 pages of international news could be handled in this fashon: it is possible that users will be interested in glancing through all of tbose pages. At present, this technique is used by almost all teletext operators. By typing one frame number the user could be presented with 15 pages over a &minute period, allowing relaxed reading of all the news presented. The systematic application of the "magazine" idea enables some 1000 pages of information to be accommodated in the (roughly) 100 time-slots available, thus increasing the usefulness of 1WN videotex tremendously.

There are, however, some drawbacks to the .'.'magazineu technique.

--

a subject which one would eipect that in Austria a t least would deserve a separate page each. The sicond problem is with the indexing of pages contained in the magazines.

Due to lack of space a t present only the title of the magazine is contained in the index. However, the pages in some magazines are rather different in nature and certainly deserve central indexing, otherwise they are lost to new or inexperienced readers. But if this information were contained in a general index, there would a the problem of access, because first the magazine has to be addressed but then the only way to get the information is to wait. Thus no direct access to the information is possible. This could be improved in principle by introducing intelligent decoders for 1WN videotex system.

Regarding the content of information, we have made comparisons between the Austrian T e l e t e z t , the West German y d e o t e z t , and the exper- imental Swiss T e l e t e z t systems. The information broadcast by the Swiss system is compiled by a consortium of major Swiss newspapers and pub- lishers, with the clear aim of producing an "electronic newspaper" with news, culture

--

even poetry

--

art, reviews, etc., although our personal impression is that the use of this medium for most of these purposes is of limited value. The West German V i d e o f e z t has a separate teletext edi- torial office in West Berlin run by the radio and

TV

companies. The main trend there is to provide a tool which basically supplements local radio and TV stations, such as by providing details of future TV and radio

(8)

programs. For example, they display frames on the detailed content of the major evening news programs Tagesschau and Heute some two hours before they are broadcast. In principle, no newspaper can compete with such a service, and teletext is also not duplicating information available from another medium. Furthermore, the "subtitling" program ( t h e ] importance of which will be explained below) is also quite advanced. The

'

Austrian Teletezt program is a "mixture" of the Swiss and the West Ger- man system, and is run by the national radio and TV authority of Austria, ORF. In addition to the above services the ORF teletext offers an interesting on-line service on four pages on the status of arrivals and departures of all flights to and from Vienna airport, Schwechat, with only a 3-5 minute delay in the latest information from the Schwechat control tower.

According to ORF data the "main headlines" on page 171 are brought up to date 50 times per day and the ORF team updates daily 1000 frames of information. According to a sample taken on January 29, 1982 a t 4 p.m.

the Austrian system broadcast 82 different pages, identifiable with separate page numbers. Out of these, 53 frames were repeated in each broadcast cycle, and 29 carried a total of 150 frames in "magazine"

fashion

-

a s will be described in greater length below. The ORF service is subdivided in five major subject categories called "registers":

Reg. 1. (Service Aktuell) contains general daily information of broad interest such as weather, exchange rates, air traffic, major events, traffic conditions, snow conditions, etc. In total, 21 pages (eight in "fixed" and 15 in "magazine" mode containing 59 frames).

Reg. 2. (Lebenshilfe) contains information on emergency telephone numbers for diverse cases (hospital, pharmacies, etc.), colysu- mer advice, general information for the handicapped (27 frames!) and a language training course (nine frames). A total of 10 pages (three "fixed" and seven in "magazine" mode with 48 frames).

Reg. 3.

(ORF-

P r o g ~ a m ) contains information on future radio and

TV

programs and a separate frame (No. 150) for carrying subti- tles for subtitled programs. In total 17 pages of information are broadcast, 14 in "fixed" and three in "magazine" mode with seven frames.

Reg. 4. ( h t e r h a l t u n g ) contains five frames of information in "magazine" fashion with eight frames with entertainment such as a chess corner and a zodiac.

Reg. 5 . (Nachrichten) contains news frames on national and international news, sport, the economy, etc. In total 21 pages, 20 of which with "fixed" information and one page with 20 (!) frames on all Austrian ski champions in "magazine" mode.

In addition, there are some further information frames which do not fall into the above categories; some of them are "service" frames of teletext, such as index pages or page 189 which contains eight frames of news in English taken from the BBC's Ceefaz, London. Thus, as mentioned above, the Austrian teletext system is a mixture of different services on 100 frames.

(9)

In the early days of videotex (1W and 2W it was considered a no-no to

"overload" pages by putting too much on them) the readability of pages was deemed to be crucially important [ I ] . Like in many situations such a dogmatic view has turned out to be wrong: although it is true that those pages which are supposed to be read (i.e., those with "actual information") should be readable and thus not overloaded, pages which are not to be read but just to be glanced a t (i.e., routing pages) could and should contain densely packed information to avoid too many routing accesses.

This philosophy of concentrating as much routing information as possible onto a page has improved the usability of videotex considerably. This becomes apparent by examining the alphabetic index of Austria's 1W or 2W videotex, or of Meyer's encyclopedia in the 2W videotex system in West Germany.

Another important lesson which is gradually being learned in connection with 1WN videotex is that, as mentioned earlier, it should not be used as "electronic newspaper" (reading of lengthy material on a

TV

is not satisfactory due to the poor quality and the lack of portability of the display; see ^[5],[6]) but should be used for up-to-date information of wide interest, for special interest groups who have to rely more on reading than other groups (such as deaf) and, in particular, in connection with ordinary

TV

programs. The use if 1WN videotex for subtitling (as is gradually being introduced in a number of countries) is an ideal example.

Although subtitling is currently only used in programs for the deaf i t is feasible, and will hopefully be pursued in the future, to use it for translat- ing interviews: the interview would carry the speakers' and the inter- viewers' words in a foreign language on the audio channel, and the subti- tle would condense the translation in the viewers' language.

1WN videotex should also prove helpful for all kinds of semi- emergency information which would presumably interest such a wide segment of the population a s t o overload any two-way videotex service, should it be offered by them.

A number of major improvements in 1WN videotex will be made possible as terminal technology develops. As was pointed out in Section 2, it is quite realistic to assume that intelligent terminals with local storage capabilities of 20-50 pages will become widely available within the next 10 years. Such terminals will upgrade 1WN videotex in a t least two ways:

they will allow alphabetic searching and the local storage of frames. For example, a user would be able t o type in an arbitrary alphabetic keyword which would be searched for by the terminal's microprocessor (either by index pages provided, or else by searching through all the pages being received), finally resulting in t h e display of all frames relevant to the specific keyword. Another application of such a terminal is to allow the user to type in the page number of a magazine (in the sense mentioned above) which is then fetched (and kept up-to-date) for later convenient and fast retrieval. The use of such terminals might permit the expansion of magazines to even more pages, giving

1WN

videotex the potential of an attractive 2000-4000 pages of up-to-date information.

(10)

Another way of increasing t h e information capacity is reported by Tydeman [2], who mentions t h a t in some of the US teletext trials different magazines are put on the systems a t chfferent, predefined times. We feel t h a t 1WN videotex will also gain further importance if it is combined with 2W videotex, as will be explained in Section 6.

One of the most significant improvements of 1WN videotex will, however, be made by a t t a c h n g printers to videotex terminals. Although t h e piece of such printers will drop to well below US $200-300 in the near future, we believe that the price is not of critical importance. In a n apartment buildings, for example, a printer could be shared ( a s is some- times done with laundry machines) between dozens of apartments, mak- ing even sophsticated printers feasible. Similarly, coin-operated printers could be made available in public places (as photocopying equipment is today), etc. The availability of such printers will make facsimile newspapers, distributed via 1WN videotex, a very attractive alternative, solving the increasingly tedious and expensive problem of newspaper delivery.

With new terminal technology [7,8,9] even t h e delivery of reasonable quality pictures (requiring about t e n times longer for transmission t h a n ordinary text frames) is possible.

The viability of the above notion is demonstrated by t h e following calculation based on the situation in Austria: suppose 1WN videotex is used for transmission of facsimile newspapers during the off-time of Austrian

TV,

e r g . for the five hours 1:OO-6:00 a.m. A t four pages per second, over 70,000 pages can be transmitted. Assuming 70 participating newspapers, 1000 frames (equivalent to more than 40 large newspaper pages and 40 pictures) a r e available for each newspaper, clearly more t h a n ample room. Observe that a printer which is supposed to print more t h a n one copy of a paper (e.g. for more t h a n one family in a n apartment building) has to use some local storage and has t o print t h e desired number of copies of e a c h page, before continuing to print t h e next page. (This will place a limit on the number of families sharing a printer).

Finally, it should be made clear t h a t t h e notion of so-called "rnulti- time" paper as speculated i n [lo] would add still a further dimension to facsimile newspaper delivery.

4.

THE

IMPACT OF 1WW

(ONE-WAY

WIDE-BAND)

VIDEOTEX, THE USE

OF CABLE 'IT,

AND

DIRECT-BROADCAST SATELUTES

1WW

videotex can be transmitted either via a dedicated broadcast channel, or via a separate CATV channel. To the authors' knowledge t h e only major experiments conducted in this direction all use CATV. Since this alternative is particularly attractive in a country with u h CATV penetration it is not surprising t h a t Canada (where over 50% of households a r e already equipped with CATV) leads in this area.

Cable TV is most suitable for this type of application since on average it can carry up t o about 40 different channels of TV and hifi-audio programs. One major problem is t o actually "fill-out" t h e capacity of CATV networks with high-quality

TV

or audio programs or, a s in our case, with other useful applications. For example, a t the time of writing, the local cable TV company of Vienna, Telekabel, provided only six different

(11)

programs (two Austrian, three West German, and one Swiss) and 16 audio programs (Austrian and West German). This service -- apart from those areas where the quality of local

TV

reception is poor -- only provides a real new service in bringing to Vienna four foreign TV channels and a few Bavarian radio channels. If, however, one takes into consideration that both Switzerland and West Germany are planning to launch their direct- broadcast satellite with Swiss and German

TV

programs around the second half of the 1980s, then the local cable

TV

company will not really bring in new services to the Vienna area any more

--

apart from the

"resource sharing" effect of the central dish, antenna to the direct- broadcast satellites, and the necessary frequency converter. 1n the long term, 'however, when the broadcast power and used frequency band of direct-broadcast satellites increase, then the size and cost of the dishes will come down to enable them to be installed on the roofs of houses, if desired. Therefore cable

TV

companies have to look for additional novel services t o attract customers. We believe t h a t 1WW videotex services (preferably using several channels) belong to this later category.

TV

companies (e.g., RTL in Luxem- bourg and Tele-Monte-Carlo in Monaco) which plan to broadcast commer- cial programs and especially advertisements for audiences in neighboring countries.

RTL,

for instance, would be received in Lyon and Hamburg and would thus cover half of France and three-quarters of Germany. It is therefore feared in those countries that these c ommercial programs would strongly compete with the national ones.

(12)

Other countries, such as Austria and Yugoslavia, see this tool as an excellent medium for conveying national culture beyond their borders.

The Austrian direct-broadcast satellite system would be able to provide 55 million people both in Austria and neighboring countries with Austrian program. In Yugoslavia

--

a country with six republics, two autonomous regions, and several languages -- there are expectations that such a sys- t e m would create closer cultural links between t h e republics and regions, and a t the same time improve the regional service. There are still a few countries in Europe where, because of geographical difficulties, the terrestrial networks could not be completed yet. A typical example is Nor- way, where t h s new technology could lead to considerable savings both in time and resources. Also, a third Austrian and Swiss program -- if the time comes to build it up

--

would be better implemented using this technology. According to [31], for example, for Norway to achieve a 95% coverage of the population it would require 154 VHS transmitters and 1,000 repeaters. In addition (due to the severe climatic conditions) the annual operation cost of the network would be of the order of US $65 million. In t h e UK, where geographical conditions are far more favorable for terrestrial. systems the annual operating costs would be about US $10 million per TV channel. In Italy and in France, it-would be around US $15 million per year, respectively. According to [31],

The annual cost per channel for a satellite based operation network with five channel satellites would be in the order of eight to nine million US dollars. The system wold consist of two satellites in orbit a t any given time (one Active satellite, one spare) and related ground facilities ( a telecommand telemetry station and the

TV

transmitting station). till the elements of the system would be insured against failures. The annual cost per channel could increase t o about 12 million dollars in the case of the most power demanding mission and on the contrary decrease to about 6 million dollars for smaller coverages.

A five channel satellite providing coverage over France or Italy for instance would lead to a n annual network cost reduction per channel in the order of 6 million dollars, i.e., about 40% reduction over the classical terrestrial network and allowing a t the same time a near 100% coverage. In the c'ase of smaller coverage area as for Britain and Germany, the percentage reduction will be lower. It would, on the contrary, be m u c h more for countries specially difficult to cover by terrestrial systems as, for instance, Norway, where the savings in operation would be about 54 million US dollars per year. Thus, broadcast satellite systems seems to be cheaper from the operational point of view.

As to the minimum initial capital investment cost of a direct- broadcast satellite System, according to [32], the following components have t o be taken into consideration:

(a) Space segment (satellite, launching, insurance premiums).

(13)

(b) Earth segment (up-link transmitters, down-link receivers (dishes)).

Some typical cost elements for present and future direct-broadcast satellite system are shown in Table 1 .

As to the cost of up-link transmitters and down-link receivers, the higher the frequency used, the more costly the up-link transmitters. Up-links in the more preferable 14 GHz band costs are about US 5500,000 a t present for a transformer and transmitter, i.e., twice as much as for a similar equipment in the 6 GHz band. It is expected that these costs will drop as time passes, and will level off a t around US $50,000. Down-link receivers, on the other hand, get smaller and cheaper with increasing frequency and power used by the satellite, prices depending on the system and quantity.

Disks range from US 5100 to 2,000 for the 12-14 GHz frequency band. As to the cost of the satellite, the higher the frequency band used and the broadcasting power, the more expensive the satellite and its launch.

Nonetheless, for direct-broadcast satellite systems, where only one or two satellites and appropriate up-link transmitters servicing a large number of users with down-link receivers are employed, it is more economical to use as high frequencies as possible. This philosophy was also adopted by the ITU and World Administrative Ra&o Conference (WARC) which allocated in 1977 higher up-link and down-link frequencies in the 12-14 GHz band for direct broadcast purposes.

In any discussion of the role of direct-broadcast satellite systems a key issue is the problem of program content. This is a crucial question, particularly for countries which already operate TV programs on well developed terrestrial networks because an entirely new program for the sake of a direct-broadcast satellite System seems to be wasteful and too expensive. According to [32], 10 hours per day of general TV programs costs US 836.5 million per year to produce (CCIR Data Report, Interna- tional Radio Consultative Committee).

According to the present allocation of WARC frequencies in Europe, Africa, and Asia, each country is allocated five specific channels (with 27

MHz

bandwidth) for direct-broadcast purposes. A t first t h s certainly brings the problem of economic utilization: a country such as Austria, for example, will find it difficult to fill its allocated channels with useful programs. However, the more channels are used, the lower will be the cost per channel for a satellite which can easily accommodate a few separate TV and radio channels. Therefore, in order to utilize fully the capacities of a modern direct-broadcast satellite, either more than one country will have to launch a common satellite for their own purposes; or one country has to utilize its channel differently. The way we a r e p r o p o s i n g is t o u s e t h e "free" d i r e c t - b r o a d c a s t c h a n n e l s f o r 1 W W v i d e o t e z purposes. With such a service a reasonable amount of information frames can be ration- ally provided, as mentioned above, which can be used quite efficiently in a future information-oriented society. The use of direct broadcast satellites for 1WW videotex will, in the long run, enable (by increasing the used transmission power of the satellite and the used frequency in the region of 20-30 GHz) that small mobile dishes and mobile terminals could be used for 1WW videotex purposes.

(14)

1.a. Capital costs of present and planned DBS systems 1.b Satelllte cost breakdown Total Satellites* Launch** Insurance Design Satellje NO. of Receiver

System Cost (Date) Premlum Llfe Mass transponders costs

(Srn) (Srn each) (Sm each) (years)

Component Deslgn and test*

lTC earth faclllty Multiplexers Orbltal positlonlng

equipment**

Antennae Solar panels and

related power englneerlng equlprnent

Cost. Sm 20 2 1.2 Average DBS 171 25

System 30 2 8 200-800 kg 214 channels

(Q=3) $200-500

ANlK B 34 28

(Canada) (Q=1) (1978) 7 400 kg ti channels' $600

ANlK C 22.5 28

,

^{522 kE} 10 channels $3300

(Canada) 165 (Q=3) (1881) 32 TV (Q= 100)

60 13

Bs-1 (Japan) 115 (Q=2) (1978) 2 352 kg 2 s 100-200

(Q=100.000)

CTS 10 350 kg ₂ S15.000(10ft dlsh)

(USICanada) 82.7 72'7 (1975) $23,000(1511 dish)

RA-6 (US) 190 1.358 kg ti

(2.6 CHz) $1-6.000

Includes dlgital T&C systems, some of power and cornrnunlcatlons

electronics;

** Thrusters and stablllzers TDF-1 .TV-SAT 30 160

(France.

Germany) NORDSAT

(Scandlnavia) 370 , ¹³

2 DBS (Q=5.000)

1.c Feasible launch vehlcles and prlces ARABSAT (ATU) 240

ECS / OTS (ESA)

(MAT 260

-PLANNING)

$200 Vehicle Cost (Sm) Orbit

CaDacttr

$400 Delta 3914 20 (kg)

-

907 Atlas-Centaur 40 1783 N-Rocket(Japan) 75 750

Artanne 1 15 4840

Arlanne 3 11-15 4000-1 1000 Shuttle(US) 11-14 4000-1 1000 2 DBS

(2.6 GHz)

SATCOL (Colurnbla) CONDOR (Andean Nat.)

1 DBS 12

8 channels $200-400 COMSAT

DBS

see table 1.b

** see table 1.c

Table 1. Typical cost of direct-broadcast satellite system components [32]

(15)

The amount of information which can be offered with 1WW videotex is about 500 times larger than with 1WN videotex, since the full bandwidth of a TV channel (6-8 MHZ) can be utilized for teletext. Thus, offering 50,000 pages of information (or more, using a variation of the "magazine" idea) is no problem with 1WW. Since the 2W videotex databases currently in use do not offer more pages t h a n of the same order of magnitude, 1WW videotex will indeed be a rival of 2W videotex a s long as 2W videotex is seen mainly as a n information service, where the amount of information is large but not super large. However, considering 2W videotex in this way implies a basic misunderstanding of the t r u e possibilities of 2W videotex.

I t is only because of this widespread misunderstanding which came about because of the ill-conceived Prestel experiment in the UK (where 2W videotex has been used primarily as information service since 1979) that it is often assumed t h a t 1WW and 2W videotex are rivals. They a r e not;

1WW will be vastly superior in performance/cost ratio to 2W videotex (assuming a n appropriate CATV infrastructure) in the area of providing large, but not huge, amounts of information which will appeal t o a sufficiently large segment of the population or to some extent if no infrastructure (e.g., a telephone network) for 2W videotex exist. Two way videotex is a reasonable alternative t o 1WW if a n adequate infrastructure for 1WW videotex is not or cannot be made available (for example, if all channels a r e already occupied with TV programs). If a n appropriate 2W videotex infrastructure is available, i t should be used for large amounts of d a t a which a r e accessed only moderately often and, most important of all, for all those myriads of applications where the interaction provided by 2W videotex is essential. The above line of thought will elaborated in depth in Section 7 , where the penetration of videotex systems and their components such a s , t h e telecommunication infrastructure a r e discussed.

Regarding the terminals t o be used for 1WW videotex preferably a n intelligent videotex decoder should be used, the one similar t o what has been referred earlier for 1WN videotex sysrems.

5. THE

ROLE

OF ZW (TWCFWAY) VIDEOTEX

In the initial period of developed 2W videotex development it was often claimed t h a t its main advantage over 1W videotex was the facility t o store "unlimited amounts" of information. Although 2W videotex does allow t h e storage of arbitrary large amounts of data ( a fact not only necessary for certain types of information such a s nationwide phone directories or large encyclopedias, but also of "philosophical" irnpor- tance, since it eliminates the need to "select" information, i.e., to exert some kind of censorship), its main importance lies not in its capability a s a n information service, but its transactional and communication potential. Before going into detail on the latter points it is worthwhle mentioning that even in t h e information providing sense 2W videotex provides potentials not possible with lW, or even

1WW

videotex.

One such instance is t h e electronic telephone directory. Even in a small country s u c h as Austria such a directory would have some 2 million entries, requiring a n estimated 300,000 videotex frames. This is beyond t h e capabilities of 1WW videotex.

(16)

Another example is the idea of offering a fully fledged encyclopedia via videotex. Ignoring the fact that other technologies such as videodiscs [ l o ] may be preferable in such instances, and assuming that modern videotex systems can handle good quality pictures (as is the case in systems such as Telidon [7], AT+T [B], Picture- FresteL and MUPID [9], hence obviating the need for locally stored pictures as proposed in [ l l ] ) , a 24- volume encyclopedia would require some 240,000 videotex frames for text, and another 160,000 frames for pictures. As in the case of an electronic nationwide phone directory or electronic super-directory as proposed in [12], such amounts of information cannot be handled by 1WW videotex. They can be handled by 2 W videotex and indeed in a commer- cially viable way, provided a sufficiently large segment of the population participates. Consider, for example, the' situation in West Germany: by 1990, some 5 million 2W videotex customers are predicted. Assuming that 2% of all videotex users are willing to subscribe to an electronic encyclopedia for a fee of US $50 per year ( t h s would compare favorably with the US $1000 buying price for each of the two major German encyclopedias Meyer and Brockhaus), this would amount to a total revenue of US $5 million per year. Assuming US $2.50 storage charge per page per year (corresponding to the current charges in Austria), $1 million would be required for storage. Based on current experiments, a staff of about 40 would be sufficient to keep the encyclopedia up to date. A t a cost of $

30,000 per man year, $2.6 million per year would remain for the acquisi- tion of material, overheads, and earnings.

Despite the fact that we believe that 2W videotex might be a viable alternative for super-large sets of information (provided the number of users is sufficiently large) we would like to emphasize that t h s is not most important aspect. (As a matter of fact, due to other more attractive alternatives videotex may never be used this way a t all [lo].) The importance of 2W videotex lies its transactional and communication capability. It would be repetitive to list once more the abundance of potential applications of 2W videotex described in many recent studies, e.g., [6], [15]; rather, we prefer to make a number of general remarks on what we consider most important.

First of all, we would like to clarify some 2W videotex applications whose significance is often overlooked: we want to distinguish between

"answer-type" and "interactive-type" systems. Classical transactional applications of videotex are often only of the "answer type." For example, when teleshopping, the user looks up some goods in a catalogue and then fills out an order form which is sent (electronically) to the information provider. No further interaction takes place between user and information provider, and hence no on-line connection between the two is necessary (this is the reason why for teleshopping PresteL-type answer pages are quite satisfactory; access to the information provider's computer is more a luxury than a necessity). Note that even telebanking, the often mentioned standard example of the need for on-line connection between user and information provider, is really of the "answer type": the user retrieves the current status of his account (which could well be done even in 1W mode, by assigning to each account one frame in the videotex computer which is only accessible to authorized users) or carries out fund transfers by filling out an appropriate form. In both cases, contrary to

(17)

often heard views, no on-line connection between user and information provider is necessary. Thus, the much emphasized on-line dialog between the user and the bank's computer is not essential for such basic applications. The only component w h c h is essential is an efficient message service.

We do not want to claim that videotex networking with interactive access to third-party computers is not important. It definitely is, for certain kinds of truly interactive applications to which we will turn below.

What we do want to claim is that for many applications for which direct access to the information provider's computer is often considered necessary considerably less is sufficient: an efficient and safe message service D21.

Indeed, we would like to claim that the value of message services in videotex is grossly underestimated by most people. Such services can be used for transactions (as explained above); for electronic mail; for communication with the deaf; for teleplaying [13]; and for a number of other applications, which form the basis of an important and still virtually unex- ploited notion of resource sharing via videotex. We would like to ela- borate briefly on the idea of resource sharing by means of an example:

suppose a user of videotex wants a number of frames of videotex in printed form, or as bgh-quality slides, or the like. Despite t h e fact that he may not have adequate equipment himself, he can request the desired output via the message system from a company offering that kind of service.

We would like to turn our attention now to truly "interactive-type"

applications of 2W videotex. Despite the fact that such applications a r e virtually non-existent in 2W videotex systems a t present, we believe t h a t they will be developed rapidly and gain increasing importance.

In a truly "interactive-type" application, on-line access t o a third- party computer is essential. This is the case where such a computer is used to r u n a sophisticated program, a game program, or to perform a computation [13]. It is also important in booking situations when rapid confirmation is required. It will play a prominent role in all kinds of teaching applications of videotex, when the instantaneous evaluation of t h e student's input is essential, and in applications where a third-party computer is used to verify the user's input (e.g, by checking his spelling).

A host of other applications is clear to anyone who just cares t o t h n k of all the ways in which we use computer terminals interactively through computer networks today. After all, 2W videotex terminals are exactly that: inexpensive, simple computer terminals whch will eventually permit access to large computer networks. It is this fact which makes 2W videotex significant beyond what 1WN videotex can ever achieve. With 2W videotex the notion of omni-present access to computer networks is slowly turning from fiction into reality.

We conclude this section by mentioning that 2W videotex will gain much by the introduction of intelligent terminals: not only will these allow more convenient searching procedures (e.g., permitting access by alphabetic keyboards and by "relational queries" in terms of a request form 1171) and the local storage of information, the notion of down- loading and executing software [13], so called "telesoftware," will also

(18)

- 1 5 -

increase the flexibility of 2W videotex tremendously.

6. THE COMBINATION OF

1W

AND 2W VIDEOTEX

In the preceding sections we have argued that each of l W N , 1WW and 2W videotex are best suited for some applications, but not for all; hence the future coexistence of all three varieties seems a definite possibility.

Assuming the widespread penetration of intelligent videotex terminals such coexistence may well turn into a fruitful cooperation, a notion first mentioned in [16]. The processing power and local storage of a n intelligent terminal may greatly increase the attractiveness of 1W videotex (see Sections 3 and 4), while some applications may be "split" between 1W and 2W videotex. For example, consider the currency exchange services offered in West Germany via third-party computers: the user accesses a bank offering such service via the gateway of videotex; he enters two currencies A and B and a n amount m; the bank's computer now computes the equivalent of m units of currency A in currency B; and sends the result n back to the user.

In a way, such applications are, from a long-term point of view, abuses of the gateway notion rather than reasonable applications thereof. Such trivial computations should be carried out in t h e user's intelligent terminal rather than overburdening (by thousands of simple requests) a third-party computer. A typical scenario of the future for such a n application could be: the user down-loads a (short) program for such currency exchange calculations from 2W videotex; this program fetches one frame of current exchange rates from 1W videotex (where such information is?offered anyway) and then performs the'desired calculations. Rather than going through 2W videotex gateways and performing some calculation i n a third party computer, one page each is retrieved from 2W and 1W videotex, much reducing t h e load on the 2W videotex system and the third party computer.

From a technical point of view the combination of 1WW and 2W videotex could and should be carried much further except that, in some countries, developments have perhaps already gone too far to make such solutions likely.

The second-generation videotex network which will go into operation in West Germany towards the end of 1983 is based on the assumption that 10% of the information is requested 90% of the time. Hence i t seems feasible to use many comparatively small regional videotex centers (for only the 10% of frequently requested information) and a single large center which will send additional frames to the regional centers as requested. Observe that the function of the regional centers could be further reduced (thus decreasing the cost of each of the large number of such centers) if that same 10% of information is distributed via 1WW videotex. The user would not even be aware of t h s fact: his intelligent terminal would first check for the required information in t h e 1WW videotex system only when failing retrieve it from 2W videotex. Consider- ing that 1WW videotex usually depends on the ability of a free

TV

channel on terrestrial networks or direct-broadcast satellite sys tems, or on the

(19)

state of the CATV network, the above proposal will only be meaningful in countries where this is available or the CATV penetration is high prior to the introduction of nationwide 2W videotex. T h s latter point is perhaps not true in some West European countries, but may apply especially to North America, t h e socialist countries, and the developing world.

The combination of 1W and 2W videotex systems in a given country (as will be shown in the following section) is not only a technical possibility but also a necessity. Thus a country with an underdeveloped telephone infrastructure w h c h is one of the present carriers of 2W videotex has to put its videotex services on 1WW broadcast videotex to the max- imum extent possible since it is quicker, easier, and cheaper to develop a necessary infrastructure based on a broadcast medium, as will be explained below.

In F ~ g u r e 1 we have compiled the presently known main videotex application classes and show which application should be mostly supported by l W N , lWW, and 2W videotex systems. In our example, we have made the assumption t h a t all these services a r e equally available to all users.

(normal (full channel (viewdata) teletext) teletext)

Information retrieval -very high simultaneous

request by users X X

-high simultaneous request X

-medium simultaneous

request

X

-low simultaneous request -information related

to TV broadcast

X

( e . g . , upcoming programs)

Games / entertainment

-games without interactivity X

-games with interactivity X

-downloading of game programs X

-promotion of

TV

_broadcast

( e . ~ . , subs titlinp)

X

Transactions/ teleshopping

-financial inf ormation X X

-advertisements X

-classified adds X

-sale catalogues X

-online ordering X

-banking transaction X

Electronic messaging

-important instant broadcast

messages

X X

-general broadcast messages

of broad interests

X

-group messages X

-individual messages X

-voting X

Data processing

-downloading of computer

programs

X X

-access to computers with

time sharing service for X

computation

-storage of user data X

Telemonitoring /home management

-emergency messages X X

-fire, burgler, medic a1 alarm

X

Education

-most frequent educational

courses

X

--specialized educational courses X

-test, examination X

Figure 1 Suitable videotex applications fow 1W and 2W videotex systems

(21)

-- penetration of host, switching and gateway computers for

"providing" and "channeling" various videotex services;

-

penetration of the different videotex service applications offered.

(b) Historical and Economic Aspects

Historical aspects concerning the state of telecommunications, broadcasting and computer infrastructures play a very important role not looked a t in depth so far. For example, a developing country with virtually no telephone infrastructure but mod- est TV broadcast facilities may to build its videotex services primarily in a 1WW videotex fashion within a reasonable time hor- izon. Economic aspects such as consumer spen&ng patterns play another important role in determining the pace of market penetration both into offices and into homes, not to mention other likely categories of videotex applications.

Fdeotex systems can be harmonized a s well. Although t h e above-mentioned topics on human, social, and legal aspects are of great importance, they will not be dealt with in this paper.

Therefore, we will only concentrate on technological, historical, and economic aspects in what follows.

_and2 W Videotex

Videotex systems (1W and 2W) as we presently know them are based on different telecommunication infrastructures. 1WN videotex is pig- gybacked on the broadcast facilities of the TV, but one could imagine that it could also be placed on a dedicated radio channel. At present, 1WW videotex systems are only in their early infancy, only one or two North American cable TV experiments along t h s line are known to us.

Nevertheless, market penetration of cable systems is an interesting phenomenon in t h s connection. 2W videotex systems as we presently know them are based on the telephone and packet-switchng computer networks, and whereas not too many historical statistics exist about market penetration of packet-switching computer networks, extensive and fascinating data are available on the telephone network and its usage.

From this, we believe that useful results can be drawn to predict the market penetration of 2W videotex networks, especially from the telecommunications point of view.

Now let us go into the details of some characteristic statistics and curves. In Figure 2 penetration curves of different media such as radio, telephone, black and white

TV,

color

TV

and cable

TV

into US households are shown. The US figures are significant also from the point of view that they represent a country in the forefront of technical development, and in many respects countries with a "time lag" between development and market penetration can make a sample for the potential way of their own domestic development. Figure 2 shows the following interesting trends.

Although the USA a t present has one of the most developed telephone infrastructures, full market penetration of US households with telephones (we have regarded 80% as full penetration) has taken about 72 years, a surprisingly long period. I t can be seen that the development of the telephone infrastructure is linearly proportional to the "richness" of the country. Figure 3 reflects the well known fact that the number of telephones per 1000 population is closely related to the GNP per capita of any country. The G N P figures in the diagrams are expressed in constant 1958 US dollars in order to exclude the effects of inflation on the curve. As it can be seen, all data follow a single diagonal line: the developing countries with low monetary resources and thus limited telephone infrastructure are a t the lower end of the curve. Austria and Japan lie somewhere in the middle, whereas the richest countries such as the USA, Canada, Sweden, and Switzerland are a t the upper end of the curve. Surprisingly, some rich Arab countries are still not within this group, due to the fact that in order to have a widespread, well developed telephone network of US standard it is not enough to be rich, but one has to be rich for a long time, without any disturbance, such as a war*.

Along the same lines, Figure 4 which shows the development curve of the telephone net- work for some selected countries, provides more evidence of this.

(24)

Figure 2 Penetration of households with telephone, radio, TV sets in the USA (%) [23, 241

(25)

No. of telephones/1000 popul.

CANADA (1977) 1970~-'

JAPAN (1977)

- - -

^--

,f

, F ~ ! ' J N ~ N D (19771

.

^{UK (1977)}

NETHERLANDS

(1977).

'AUSTRALIA (1977) _-

.

FRG (1977)-

s FRANCE (1977)

ITALY (1977)

'~BELG~UM

⁽¹⁹⁷⁷⁾

.

B HUNGARY (1977)

Figure 3 Number of telephone per 1,000 population in 1977 for selected countries and for the USA between 1890 and 1977 [22, 23,261